Hi, my name is Krista Swisher, and I am a neurointensivist and clinical neurophysiologist at Atrium Health Carolinas Medical Center in Charlotte, North Carolina. Today, I will be speaking about the diagnosis and management of seizures and status epilepticus in the ICU patient population. As for my disclosures, I've received both speaker's honorarium and advisory board fees from UCB. In addition, I am on the DSMB for a trial being performed by Marinus. Our learning objectives for today are as follows. We first need to identify how seizures present in critically ill patients, so we know how to recognize them. Then we need to appreciate the impact of seizures on neuronal injury and outcomes. Why is it so important that we appropriately treat them? And lastly, we will identify the guideline-based tiered approach to the treatment of seizures and status epilepticus in our patients that we see in the ICU. The definition of convulsive status has evolved over time based on new data about the harmful effects it has. And it's generally accepted that convulsive status epilepticus is defined as continuous tonic-clonic seizures for at least five minutes or two or more seizures without recovery in between. The American Clinical Neurophysiology Society defined non-convulsive status as electrographic seizure activity for 10 or more minutes or seizure activity for 20% or more of any 60-minute period of recording. Now, the Neurocritical Care Society developed a unifying diagnosis of both convulsive and non-convulsive status that is as follows, five or more minutes of clinical and or electrographic seizure activity or recurrent seizure activity without recovery of consciousness in between. Now, there's an electrographic criteria for the diagnosis of non-convulsive status, and this is important for you to understand as you're communicating with a neurophysiologist. This includes patients that have epileptic discharges that are more frequent than two and a half hertz, so more than 2.5 per second, or if these discharges are less than that frequency, but there's rhythmic delta or theta activity and it's associated with one of the following criteria. The one subtle clinical ictal phenomenon that you can see here on the right, such as nystagmus or gaze deviation, two, spatial temporal evolution, so evolving over space or time, and three, EEG and clinical improvement after an IV anti-seizure medication administration. So if patients fail initial treatment and they progress in terms of their seizure activity, they now are categorized as having refractory status epilepticus or RSE. That's defined as clinical or electrographic seizures that persist despite adequate doses of an initial benzodiazepine and an anti-seizure medication. Then if seizures continue beyond that, they are defined as having a super refractory status epilepticus or SRSE. And that is seizures that continue beyond 24 hours after the initiation of anesthetic therapy, or if they recur during the anesthetic weaning. And these definitions apply both to convulsive and non-convulsive seizures. So it's important to understand how an electrographic seizure is identified because as you receive these reports from the neurophysiology team, when you have a deeper understanding, you'll be able to provide better clinical care. And so an electrographic seizure is defined as any pattern lasting at least 10 seconds and having repetitive, generalized, or focal epileptiform discharges averaging at least 2 1⁄2 hertz. So 2 1⁄2 of these per second. Or any rhythmic or periodic pattern less than 2 1⁄2 hertz, but it has definitive evolution. And so the example shown below here is an example of definitive evolution in terms of frequency. As you can see, it starts out faster at 5 hertz, and then it changes to 3 hertz, and then it changes to 2 hertz. So this is an example of a left hemispheric electrographic seizure. It appears to be starting in the left hemisphere, probably in the left temporal region. And by the yellow arrows, you can start to see that there's rhythmic, sharp activity, and that gradually increases in terms of frequency. And it gradually changes in terms of the morphology as well. This is the subsequent EEG page. And in the left temporal region and also in the left central and parietal regions, you can see those rhythmic spikes that are gradually becoming more and more apparent. Status is something that we'll come across fairly frequently in the ICU. And as many as 20% of patients in the ICU will have seizures in status. And for patients with a status, 12 to 43% will develop refractory status, and 10 to 15% will develop super refractory status. The case fatality rate is 15%, and this is higher in elderly patients and in patients with refractory status. The International League Against Epilepsy has defined seizures as being due to known causes or symptomatic or unknown causes, which are cryptogenic. So known causes, things like stroke, ICH, TBI, et cetera, or it could be remote, such as a prior stroke. For patients in the cryptogenic category, they may have new onset refractory status, also known as NORS. And this is patients with no readily identifiable cause, although autoimmune or viral etiology may later be found. And anti-NMDA encephalitis is a frequent cause of this. The etiology of status, refractory status, and super refractory status is a large determinant of that patient's mortality. The etiology status and refractory status are similar, but the patients with super refractory status are more likely to have encephalitis. Oftentimes, it's immediately apparent why our patient is seizing. For example, a patient with a left frontal ICH has left frontal seizures on EEG. But if you don't know why your patient's seizing, you need to be able to reference a table such as this to be able to go through these one by one. It's also important to remember that patients without acute neurologic illness still may have status. There was a study of comatose MICU patients that found that 10% of them were in non-convulsive status. So really, why are we so worried about status epilepticus? It's because it causes neuronal damage. And this has been demonstrated time and time again in animal studies. During convulsive status epilepticus and non-convulsive status epilepticus. Serum neuron-specific enolase, or NSE, is increased in both convulsive and non-convulsive status. And there are numerous mechanisms of cell death in seizures and status, and these include excitotoxicity, mitochondrial dysfunction, necrosis, and apoptosis. There can be MRI changes in status, and these include T2 hyperintensities, restricted diffusion, ADC changes, leptomeningeal enhancement. And these changes are most commonly seen in the cortex and the gray matter, in the basal ganglia, the thalami, the hippocampi, and the corpus callosum. The example MRI on the right is a flare sequence, and during status, you can see that there's a flare hyperintensity in the right perisylvian region, and that appears to have resolved on the subsequent follow-up imaging after status was treated successfully with pentobarbital. Continuous EEG monitoring is essential in the ICU and that's because 92% of seizures in the ICU are non-convulsive. Why can't we just get a spot EEG? Well that's because of all the patients that will have non-convulsive seizures or non-convulsive status, a 60 minute EEG will miss about half. So in a cohort of patients with seizures on EEG, a study showed that at the start of the file, 15% were seizing, at 1 hour, 56% were seizing, 6 hours, 77%, 12 hours, 82%, and at 24 hours, 88%. So you need to run the continuous EEG longer if the patient has highly epileptiform activities such as periodic discharges, potentially rhythmic delta activity, epileptiform activity or BIRDs which are also brief, intermittent, and rhythmic discharges. This is seizure-like activity but lasting less than 10 seconds. So just like with stroke where time is brain, same thing goes for seizures. In a study of subarachnoid hemorrhage patients, every hour of seizure activity increases the odds ratio by 1.1 to a three-month disability and mortality. Similarly, in patients with intracerebral hemorrhage, if they develop status epilepticus, that increase their mortality by an odds ratio of 1.18, and their morbidity increased by an odds ratio of 1.53. Next, we're gonna talk about the management of seizures and status in the ICU. And there's gonna be three pillars. First one, benzodiazepines. Second one, anti-seizure medications or ASMs. And the third, ongoing general ICU medical management. In the US, there are two sets of guidelines for the treatment of status. American Epilepsy Society, which is on the left, and the Neurocritical Care Society, which is on the right. Both follow the convention of status being defined as a seizure lasting for more than five minutes, or multiple seizures without recovery in between. Of note, the NCS guidelines specify it's meant for both convulsive and non-convulsive, but for the AES guidelines, those are geared towards convulsive. They both are generally similar in that they emphasize ongoing general medical care, initially benzodiazepines, followed by second line or urgent control therapy, and then if seizures are still ongoing, then followed by third phase therapy or refractory therapy. And we'll talk about this in a little bit more detail in the subsequent slides. This slide here is putting all the definitions together with time starting at the top and going towards the bottom of the slide, then the definition of what phase of status they are in as it relates to where they are in terms of their treatment. Now, I probably don't need to dive too much into this at a Neurocritical Care Society meeting, but we do need to point out that general ICU management of status patients is of utmost importance. First, ABCs, paying attention to airway, breathing, and circulation, and managing that as patient's hemodynamics. Intubation may be indicated by cardiorespiratory parameters or if a sedative anti-seizure medication infusion is needed for status management, such as propofol or midazolam. You need to have continuous vital sign monitoring. The patient should have at least two large IVs. They may or may not need fluid resuscitation. Obtain a finger stick glucose level ASAP, and always give thiamine before dextrose. And then to send a rainbow of labs, as listed here. And for patients that are female of childbearing years, it's important not to forget a urine pregnancy test. Then the patient should get a non-contrast head CT when they are stabilized. If their clinical history indicates that there may be an infection present, consider a lumbar puncture. They may need initiation of pressors if their MAP drops below 65 and they become hypotensive. Initiate continuous EEG monitoring as soon as possible, preferably even in the ED, if that is feasible at your institution, and then start working on transfer to an ICU. The table on the right shows the list of complications that can be present in generalized convulsive status epilepticus, and you need to be prepared that these will occur. So for first-line therapy, the options are IV lorazepam, IM midazolam, or IV diazepam. The dose for IV lorazepam is 0.1 mg per kg IV with a maximum of four milligrams per dose. Based on the weights of the patients that you will see, the vast majority of patients should be getting four milligrams per dose. And this is preferred if IV access is already in place. However, if the patient does not have an IV, then IM midazolam should be given because this has the most reliable IM absorption. The dose is 0.2 mg per kg with a maximum of 10 mg per dose. I'd like to also point out the RAMPART trial, and this was a trial that looked at adult and pediatric patients with status, and they randomized four milligrams IV lorazepam versus 10 milligrams IM midazolam. And they found that midazolam was superior to lorazepam, presumably because the rate at which the medication could be administered, because it did not require placement of an IV. However, both groups had similar rates of recurrent seizures and intubation between the two groups. So I talked a little bit about benzodiazepine timing and dosing, but I'd like to use this slide here to emphasize it. There was a study that was done, shown here on the left side of the screen, in pediatric patients with convulsive status, and only about 34% of them received timely benzodosing within 10 minutes. The patients that received delayed benzodosing beyond 10 minutes had an increased odds of death, longer seizure duration, increased need for IV anesthetics, increased hypotension, and the rest of their anti-seizure medications were delayed. Then a trial on the right looked at the dosing, and in a retrospective cohort of convulsive status patients, inadequate benzodosing, so less than four milligrams of lorazepam, resulted in patients being more likely to progress to refractory status and more likely to progress to non-convulsive status with coma. In the patients that received adequate benzodosing, which is four milligrams of lorazepam, that unfortunately only occurred in 31% of patients. So in second lane therapy, all patients presenting a status should receive an additional anti-seizure medication after the benzodiazepine with the goal of obtaining or maintaining seizure cessation. And all patients will fit into one of those two categories. The other goal is to rapidly achieve appropriate serum anti-seizure medication levels. And this should be given within five to 10 minutes of status onset or patient arrival. And why do we give a loading dose? The graph on the top shows that if you just started as a maintenance dose, that patient is not going to have serum levels above the minimum effective concentration level for at least a day. But the graph on the bottom shows that if you give a loading dose followed by the maintenance dosing, then the patient will have serum levels above the minimum effective concentration. So one question that's been looked at time and time again is what is the best second line therapy to initially administer? And as of recent, we finally have a good answer to this. The ESET trial, or the Established Status Epilepticus Treatment Trial, was a multi-center randomized double-blind trial that looked at levotriacetam 60 mg per kg, phosphonatone 20 mg per kg, and valproic acid 40 mg per kg. And there was no difference between the groups for the primary endpoint or adverse events. So basically you can take your pick of one of these three medications, but what is needing to be emphasized here is that they have to be appropriately dosed based on the mg per kg as shown here. Phosphonatone is preferred over phenytoin because it is able to be loaded faster, causes less local irritation, and can be given IM. Dosing is one-to-one IV to PO, but you do need to increase the dose by 33% if you're switching to via tube because it bonds to the tube feeds and to the OG or NG tube itself. Levels are going to be for total 10 to 20 and free one to 2.5. But the term therapeutic level is misleading because if the patient, say for example, has a free level of 0.8, but they're no longer seizing, then that is actually therapeutic for that patient. It is highly protein-bound, so it would have a lot of drug-drug interactions. You need to check the free levels in ICU patients because they often have low albumin or due to other drug-drug interactions. And it has a narrow therapeutic range, so side effects to look out for are somnolence, nausea, ataxia, nystagmus, and it can even become epileptogenic if it is very, very high. You can check the total and free levels one hour after the loading dose is complete. Foproate, like phenytoin, is extensively protein-bound, so will result in a lot of drug-drug interactions. Dosing IV to PO is one-to-one. The side effects are shown here, but ones that I like to point out because we take care of patients with ICH, subdurals, other bleeding issues, is that Foproate can cause thrombocytopenia, platelet dysfunction, and coagulopathy, although this is quite rare. You can develop hyperammonemia, but this typically takes time to develop and is typically asymptomatic. Deprecate levels can be obtained immediately after the loading dose. You should send a total level only because the free takes days to resolve at most institutions. Levotracetam is devoid of drug-drug interactions. The IV to PO dose adjustments is, or dosing ratio is one-to-one. Renal dose adjustment is needed for AKI and ESRD, and delirium and agitation are potential problems with this medication. So out of those three, what should you choose? For the 2020 American Epilepsy Society Guidelines Committee Review, they stated that for convulsive status, there's no consistent differences between levotracetam, phenytoin slash phosphenytoin, and valproic acid for status persisting after first-line administration of a benzodiazepine. Next, we'll talk about latosomide, and this should be a consideration in patients particularly with non-convulsive or non-convulsive status. The FDA-approved loading dose is 200 milligrams, but most studies have used 400 to 600 milligrams IV. The American Epilepsy Society recommends 10 mgs per kg loading dose, and a typical maintenance dose is 200 milligrams BID. Dose adjustments are needed in renal and hepatic insufficiency, and you need to watch out for cardiac conduction side effects. The TRENDS trial was a trial that looked at patients with non-convulsive seizures that randomized phosphenytoin, 20 mgs per kg loading, versus lacosamide, 400 milligrams IV loading dose, and it demonstrated that lacosamide was non-inferior to phosphenytoin. The rate of adverse events was similar in both groups in terms of hypotension and bradycardia. Braviracetam is a medication in the racetam class, and this has a high affinity for the SV2A receptor, which you probably remember that levatoracetam acts on the SV2A receptor, but braviracetam has a much higher affinity for this. It also has faster blood-brain penetration than levatoracetam. It's given as a two-minute bolus. The optimal loading dose is unknown, but people often use 200 milligrams. The maintenance dose is given as 100 milligrams BID, and one nice thing about this medication is that renal adjustment is not needed, but its use is not recommended in end-stage renal disease because data is limited here. The effectiveness and status is unknown. So if you've given a benzodiazepine and ASM, but status is still present, your patient has refractory status epilepticus. Now you have two options. You can load another anti-seizure medication, or you can start an IV anesthetic. There have been some studies that have shown greater complications and high risk of death in convulsive status patients that received anesthetics compared to ones that only received non-sedating anti-seizure medications, but these studies are really complicated by selection bias. The 2020 American Epilepsy Society Treatments Committee for convulsive refractory status stated that there's no strong evidence to support either treatment approach. Glucosamide 400 milligrams IV has the highest level of evidence, which is class C, when compared with IV briviracetam, ketamine, levitracetam, midazolam, pentobarbital propofol, or alproic acid, and all of these they categorize as level U, which is data inadequate. This is a table that was published by the AES Treatment Guidelines, one that I've already referenced a few times already, and really what's most important about refractory therapy is how we're going to use midazolam and propofol, but there's a variety of other refractory treatment options that are shown here in this table. Now, what you're gonna use most often for refractory therapy is midazolam and propofol drips. The dosing is different than how you're gonna use it for sedation. Midazolam drip is dosed at 0.1 mix per kick per hour and titrated up to two mix per kick per hour until you get an appropriate endpoint. Doses of over 100 milligrams per hour are common. Don't pause these for exams. It's unlikely that the degree of EG suppression is associated with outcome, so there is no way to recommend a specific birth suppression ratio, for example. Similarly, the length of anesthetic treatment is ideally is unknown, and typically we do 24 to 48 hours before weaning. In general, most neurointensivists prefer midazolam drip over propofol drip due to the risk of propofol infusion syndrome, which manifests as cardiac failure, rhabdo, metabolic acidosis, and kidney failure, and is often fatal. The risk is higher with longer exposure, higher doses than in pediatric patients. In patients on a propofol drip, follow CK, triglycerides, creatinine, lactate, and pH, and do be aware that tachyphylaxis can develop with midazolam, which is over time there is less efficacy due to GABA receptor internalization. Ketamine is often used as an adjunctive therapy in patients with refractory or supra-refractory status. This has multiple mechanisms, including NMDA antagonism. It doesn't suppress cardiovascular tone, and it doesn't cause respiratory depression. You typically start at two mix per kick per hour, and then you go up to 10 mix per kick per hour. It has a synergistic effect when used with benzodiazepines, and that's because in status, GABA receptors are downregulated, and NMDA receptors are upregulated. So this is why it can be particularly helpful. The end point is seizure suppression. So it doesn't cause burst suppression. So when you're titrating up the dose, you need to be aware of that. It's not in any guidelines, but I think it's important to talk a little bit about how I utilize it. I typically start ketamine when a VersaDrip is getting above 0.3 mix per kick per hour. And why do I start it so early? Well, one, it has very little side effects, and there's very little downside in this patient population. But two, it has been shown to be neuroprotectant against glutamate-induced neuronal necrosis in animal models of status. Things to watch out for, increased secretions and tachyarrhythmias. And previously, it was thought to cause elevated ICPs, and this has been debunked. Other treatment options for suprarefractory status that we won't discuss here are isofluorine, ketogenic diet, hypothermia, electroconvulsive therapy, transcranial magnetic stimulation, and vagal nerve stimulation. Unfortunately, we don't have time to discuss all the other ASMs that we use in this patient population. Other things that you should consider and be comfortable with are phenobarbital, Clobizam, Parampanel, Topamax, Gabapentin, and Pregabalin. And lastly, I'd like to talk about autoimmune encephalitis. Compared to status of other causes, this is more likely to be refractory and in younger patients. We've been detecting more and more of this over time as we've identified more and more cell surface autoantibodies that are causative here. Some of these are anti-GAD, anti-NMDA, and the list of these is very long. Because the turnaround times for these perineoplastic panels can be very slow, if you have a high index of suspicion, you need to start treatment as soon as possible. And this includes steroids, plasmapheresis, IVIG, and other immune modulating agents as shown here. I would like to thank Dr. Arti Sarwal and Dr. Rhonda Kadena for inviting me to give this lecture, and I'd like to thank SCCM for giving me this opportunity. And I wish everyone good luck on their boards.

seizures

status epilepticus

ICU patients

continuous EEG monitoring

benzodiazepines

autoimmune encephalitis